Functional characterization of "Bartonella" effector protein - BepE during "in vivo" and "in vitro" infection

The bartonellae is a family of gram-negative, fastidious, facultative intracellular, zoonotic bacteria. Most of the Bartonella species are highly adapted to establish asymptomatic bacteremia of their reservoir host within which the bacteria colonize erythrocytes as privileged host niche and develop long-lasting persistent infections. Bartonella uses a VirB type IV secretion system (T4SS) to translocate Bartonella effector proteins (Beps) into the infected cells. By using such a tool box it subverts host cellular functions in order to establish a safe niche for replication and survival.
This thesis aimed to elucidate the role of one of the effector proteins – BepE in the establishment of Bartonella infection by using in vivo and in vitro infection models. Started in December 2006, my primary aim was to establish a suitable model for pathogen - natural host interaction. In order to closely mimic the reservoir host infection by BartonelIa, I have adapted the rat intra-venous (i.v.) to intra-dermal (i.d.) infection model, inoculation of B. tribocorum (Btr) in the ear dermis of the animal. This route of infection reflects the natural way of Bartonella transmission by arthropods when the bacteria are inoculated in the skin of a mammal via the feces of a vector after animal scratches. The Btr wild-type i.d. infected animals developed blood stage infection, which started around 7-8 days post infection and lasted for 10 weeks. It was a long-term bacteremic infection without obvious clinical manifestations, a hallmark of the reservoir host infection by Batonellae. The time delay that Btr took to appear in blood could correspond to the way that bacteria need to pass from the derma to the lymphatic-blood system and to the possible interaction with the innate immune system. In summary, the rat i.d. model enabled us to distinguish Bartonella factors involved on two different phases of the infection: early phase, prior seeding into the blood and the blood stage. On those two stages bacteria have different environment to interact with, and assumably different strategies to cope with the host immune system. The rat i.d. infection model revealed BepE as a critical factor in the establishment of reservoir host bacteremia. The expression of BepEBtr could rescue the abacteremic phenotype of Btr ΔbepDE mutant and enabled the strain to reach the blood. Heterologous complementation of Btr ΔbepDE phenotype with BepEBhe suggests that this function of BepE is conserved between different species of Bartonellae. Even more, I could demonstrate that the C-terminal BID domains are having the specific function but putative phosphotyrosine-containing N-term of BepE does not play an essential role in the establishment of long-term bacteremic infection of the natural host by Bartonella.
Another phenotype of BepE but in vitro was observed during the infection of primary endothelial cells HUVECs with Bhe ΔbepE (and ΔbepDEF) mutant(s). Besides erythrocytes, endothelial cells represent another major target cell type for Bartonellae demonstrated as bacillary angiomatoses within the incidental host environment, mostly in immunocompromized human patients. HUVECs infected with Bhe stain that lacked BepEBhe revealed disturbed rear edge detachment during migration and followed with the fragmentation of cell body. This phenomenon was inhibited by pbepEBhe expression in Bhe ΔbepE (and ΔbepDEF) as well as, by T4SS independent expression of pbepEBhe in HUVECs by transfection prior the infection with Bhe ΔbepE (and ΔbepDEF). We found that the cell fragmentation of infected HUVECs is T4SS dependent and is a secondary effect of translocated Beps, potentially the Beps involved in the invasome formation. Further we conclude that the C-terminal BID domains of BepEBhe are sufficient to interfere with the cells fragmentation process. From this we could hypothesize that primary infected cells in i.d. infection model of rats may also undergo fragmentation or impaired migration when infected with Btr ΔbepDE and then Bartonella does not succeed to reach the blood system and colonize red blood cells.
Further, I introduced the i.d. in vivo infection of Rosa 26-loxP-egfp Balb/c mice and in vitro infection of mouse Bone Marrow-derived Dendritic Cells (BMDCs) with B. birtlesii (Bbi) strain that is expressing Cre-BID fusion protein. The in vitro model showed for the first time a Bartonella effector protein translocation in primary immune cells of the reservoir host. This finding builds a strong basis for the hypothesis that primary infected cells in vivo may be the DCs (Langerhance cells or dermal DCs) in the skin of infected animal. DCs are the sentinels of the immune system that constantly sample the environment for the “danger signal”. Thus, they represent one of the candidate cells in the derma to be targeted by Bartonella after inoculation of the bacteria from the feces of arthropod vector. Infected DCs could serve as Trojan horses to carry and disseminate Bartonella from derma to lymphatic–blood system.